Optics Letters,
Journal Year:
2024,
Volume and Issue:
49(20), P. 5886 - 5886
Published: Sept. 23, 2024
Fresnel
incoherent
correlation
holography
(FINCH)
records
coaxial
holograms
for
wide-field
3D
imaging
with
light,
but
its
temporal
phase-shifting
strategy
makes
dynamic
challenging.
Here,
we
present
a
compact,
portable
single-shot
mirrored
(SSPMS)
module
that
can
be
easily
integrated
into
the
FINCH
system,
achieving
secondary
modulation
of
self-interference
beams
to
enable
simultaneous
acquisition
four
phase-shift
in
single
exposure.
Compared
previously
reported
methods
use
diffraction
gratings
spatially
separate
at
specific
angles,
this
duplicates
laterally
shifted
beam
using
simply
modified
Michelson
interferometer,
so
obtained
via
are
free
from
optical
aberrations
or
higher-order
diffracted
light
noises.
The
feasibility
proposed
method
is
experimentally
demonstrated
through
grayscale
scenes.
Phase
recovery
(PR)
refers
to
calculating
the
phase
of
light
field
from
its
intensity
measurements.
As
exemplified
quantitative
imaging
and
coherent
diffraction
adaptive
optics,
PR
is
essential
for
reconstructing
refractive
index
distribution
or
topography
an
object
correcting
aberration
system.
In
recent
years,
deep
learning
(DL),
often
implemented
through
neural
networks,
has
provided
unprecedented
support
computational
imaging,
leading
more
efficient
solutions
various
problems.
this
review,
we
first
briefly
introduce
conventional
methods
PR.
Then,
review
how
DL
provides
following
three
stages,
namely,
pre-processing,
in-processing,
post-processing.
We
also
used
in
image
processing.
Finally,
summarize
work
provide
outlook
on
better
use
improve
reliability
efficiency
Furthermore,
present
a
live-updating
resource
(
https://github.com/kqwang/phase-recovery
)
readers
learn
about
Abstract
Incoherent
digital
holography
no
longer
requires
spatial
coherence
of
the
light
field,
breaking
through
imaging
resolution
coherent
holography.
However,
traditional
reconstruction
methods
cannot
avoid
inherent
contradiction
between
temporal
and
signal‐to‐noise
ratio,
which
is
mitigated
by
deep
learning
methods,
there
are
problems
such
as
dataset
labeling
insufficient
generalization
ability.
Here,
a
self‐calibrating
approach
with
an
untrained
network
proposed
fusing
plug‐and‐play
nonlinear
block,
forward
physics
model,
physically
enhanced
neural
network.
Measurement
consistency
total
variation
kernel
function
regularization
used
to
optimize
parameters
invert
potential
process.
The
results
show
that
method
can
achieve
high
fidelity,
dynamic,
artifact‐free
3D
using
single
hologram
without
need
for
datasets
or
labels.
In
addition,
peak
ratio
reconstructed
image
improved
factor
4.6
compared
previous
methods.
leads
considerable
performance
improvement
on
inverse
problem,
bringing
new
enlightenment
high‐precision
unsupervised
incoherent
holographic
imaging.
Optics Express,
Journal Year:
2024,
Volume and Issue:
32(6), P. 10563 - 10563
Published: Feb. 27, 2024
Fresnel
incoherent
correlation
holography
(FINCH)
enables
high-resolution
3D
imaging
of
objects
from
several
2D
holograms
under
light
and
has
many
attractive
applications
in
motionless
fluorescence
imaging.
However,
FINCH
difficulty
implementing
dynamic
scenes
since
multiple
phase-shifting
need
to
be
recorded
for
removing
the
bias
term
twin
image
reconstructed
scene,
which
requires
object
remain
static
during
this
progress.
Here,
we
propose
a
dual-channel
noncoherent
compressive
method.
First,
pair
with
π
phase
shifts
obtained
single
shot
are
used
noise.
Then,
physic-driven
sensing
(CS)
algorithm
is
achieve
twin-image-free
reconstruction.
In
addition,
analyze
reconstruction
effect
suitability
CS
two-step
phase-shift
filtering
different
complexities.
The
experimental
results
show
that
proposed
method
can
record
hologram
videos
without
sacrificing
field
view
or
resolution.
Moreover,
system
refocuses
images
at
arbitrary
depth
positions
via
computation,
hence
providing
new
fast
high-throughput
Optics Letters,
Journal Year:
2025,
Volume and Issue:
50(4), P. 1261 - 1261
Published: Jan. 22, 2025
Self-interference
digital
holography
extends
the
application
of
to
non-coherent
imaging
fields
such
as
fluorescence
and
scattered
light,
providing
a
new
solution,
best
our
knowledge,
for
wide
field
3D
low
coherence
or
partially
coherent
signals.
However,
cross
talk
information
has
always
been
an
important
factor
limiting
resolution
this
method.
The
suppression
is
complex
nonlinear
problem,
deep
learning
can
easily
obtain
its
corresponding
model
through
data-driven
methods.
in
real
experiments,
it
difficult
paired
datasets
complete
training.
Here,
we
propose
unsupervised
method
based
on
cycle-consistent
generative
adversarial
network
(CycleGAN)
self-interference
holography.
Through
introduction
saliency
constraint,
model,
named
crosstalk
suppressing
with
neural
(CS-UNN),
learn
mapping
between
two
image
domains
without
requiring
training
data
while
avoiding
distortions
content.
Experimental
analysis
shown
that
suppress
reconstructed
images
need
strategies
large
number
datasets,
effective
solution
technology.
Scientific Reports,
Journal Year:
2024,
Volume and Issue:
14(1)
Published: March 18, 2024
Scanning
electron
microscopy
(SEM)
is
a
crucial
tool
for
analyzing
submicron-scale
structures.
However,
the
attainment
of
high-quality
SEM
images
contingent
upon
high
conductivity
material
due
to
constraints
imposed
by
its
imaging
principles.
For
weakly
conductive
materials
or
structures
induced
intrinsic
properties
organic
doping,
quality
significantly
compromised,
thereby
impeding
accuracy
subsequent
structure-related
analyses.
Moreover,
unavailability
paired
high-low
in
this
context
renders
supervised-based
image
processing
methods
ineffective
addressing
challenge.
Here,
an
unsupervised
method
based
on
Cycle-consistent
Generative
Adversarial
Network
(CycleGAN)
was
proposed
enhance
samples.
The
model
can
perform
end-to-end
learning
using
unpaired
blurred
and
clear
from
well-conductive
samples,
respectively.
To
address
requirements
structure
analysis,
edge
loss
function
further
introduced
recover
finer
details
network-generated
images.
Various
quantitative
evaluations
substantiate
efficacy
improvement
with
better
performance
than
traditional
methods.
Our
framework
broadens
application
artificial
intelligence
holding
significant
implications
fields
such
as
science
restoration.
